Imagine a compliance officer at a mid-size accounting firm. She has been using an AI tool her company never approved, for eight months. Uses it to summarise contracts, flag regulatory inconsistencies, and draft initial client memos.
She does not hide it exactly, but she does not mention it in meetings either. But, she is not alone. According to new research, she is one of a third of all lawyers, accountants, and compliance professionals doing exactly the same thing right now.
This is where AI in professional services actually is in 2026. Not in the boardroom slide decks. Not in the strategy documents. But, in the gap between what firms promised and what employees needed, being quietly filled by tools no one authorised and no one is monitoring.
The Numbers Behind the Quiet Crisis
This week Thomson Reuters unveiled its 2026 Future of Professionals report, which is based on a worldwide survey of 1,800 professionals in the legal, tax, audit, accounting, compliance, risk and global trade sectors. Findings: 74 percent of professionals are already using AI tools weekly, but 91 percent feel their organization isn’t making the most out of what AI can achieve.
That AI implementation gap between usage and value is the story. Technology is not the problem. The strategy to use it properly is. 35 percent of professionals report that their company’s AI objectives don’t appear in their day to day routines. Almost 20% report that their organisation lacks a strategy altogether.
When the tool is not officially available, what do people do? They find one. 41 percent of lawyers, accountants and compliance professionals report that their firm is “slow to be going ahead with AI,” while one out of three says they are using AI tools that they believe their organisation has not approved.
The twist is that it’s jobs that depend on accountability. 96 percent of professionals state that their AI needs to protect confidential data. 94 percent require verified, authoritative content. 90 percent need outputs they can explain and defend. However, 41 percent lack access to professional-grade tools that meet these standards.
The result is a quiet catastrophe of invisible risk. Client files run through unverified tools. Confidential information fed into systems with unknown data retention policies. Outputs that look credible but carry no chain of authority. In a sector where a single wrong clause or missed regulation can trigger liability, this is not a minor inefficiency. It is a structural exposure that most senior leaders have not yet seen on their dashboards.
Talent Has Already Decided
Here is what makes this problem urgent rather than eventual. The professionals experiencing this AI implementation gap are not staying patient. One in four professionals, 24 percent, say they would consider leaving within two years if they do not see the AI value they expect. 13 percent say they are considering leaving within 12 months.
Meanwhile, the leaders responsible for fixing this think the pressure is still comfortably distant. Almost half of senior leaders believe meaningful talent pressure is still at least three years away. Three years. While one in eight of their best people is thinking about leaving within one.
62 percent of professionals say access to professional-grade AI would be a factor in accepting a new role. Among those already using it properly, nearly one in three say they would turn a role down without it. Moreover, AI access has become, for this generation of professionals, what a good pension scheme or clear promotion path was for a previous one. It is a signal about whether a firm takes its people seriously.
AI Implementation Gap: Clients Are Not Waiting Either
The external pressure is building at the same pace. 78 percent of corporate clients now consider AI-enabled quality improvements very important or essential. Yet just 6 percent say most of their providers are actually delivering it.
Six percent. That means 94 percent of clients who expect AI-driven quality from their legal and professional service providers believe they are not receiving it. And they are starting to act on that gap. Within 12 months, 32 percent will be reconsidering their provider relationships, with a third putting more than USD 1 million in annual work at risk. Combined this puts approximately USD 143 billion in US legal and accounting revenue under active reconsideration, based on AI delivery performance alone.
However, this is not a future scenario. These are decisions being made in 2026, by clients who expected their providers to have moved faster and are now shopping for alternatives.
AI Implementation Gap: The Accountability Gap
Steve Hasker, President and CEO of Thomson Reuters, framed the core issue plainly: “When outputs shape legal judgments, regulatory filings, or client advice, ‘almost right’ isn’t good enough.”
That is the distinction that separates this sector from almost every other industry navigating AI right now. A marketing team using an unverified AI tool to generate copy carries one level of risk. Also, a lawyer using the same tool to draft a regulatory filing carries a different one entirely. The liability follows the professional, not the algorithm.
The technology, as the report makes clear, is not waiting for anyone. The AI implementation gap is not in AI readiness. It is in execution. And the benchmark is now accountability.
The compliance officer running unauthorised tools on client files did not do it because she is reckless. She did it because she needed to do her job and her firm had not given her a better option. That is the gap. And it is costing professional firms more than they currently know.
Read More: Google turns Android into a powerful AI assistant with Gemini Intelligence
